The practical implementation of deep convolutional networks faces challenges such as feature redundancy and diminished learning efficiency. To address these issues, this paper focuses on Deep Convolutional Neural Networks (DCNNs), exploring two key aspects: feature extraction and training efficiency. The aim is to enhance the prediction accuracy, robustness, and generalization of deep convolutional networks. The proposed algorithms are applied to image classification. In response to the problem of feature redundancy in DCNN, an improved DCNN algorithm based on residual Dropout convolution (RD Conv) is proposed. The proposed algorithm includes residual Dropout paths and convolutional paths, which are randomly switched during training, resulting in random changes in network depth and enhancing the diversity of feature extraction. The residual path randomly selects input features, compresses and amplifies them, increasing the diversity of feature inputs in the downstream parameter layer. Only convolutional paths are used in prediction to ensure stable prediction results. To address the issue of decreased saturation training efficiency in DCNN filters caused by the fixed positions of the activation and saturation regions of the activation function, an improved DCNN algorithm based on a random activation function is proposed. By combining the different primary and secondary activation functions in the active and saturated regions, the filters in the saturated region can participate in the weight update process, increasing feature generation. Our RD Conv proposes a new dropout module that simultaneously utilizes the dropout layer, convolution layer and batch norm layer, while eliminating variance offset. Additionally, our random activation mechanism integrates two different activation functions into a new activation function, improving training efficiency. By using two linear functions as sub-activation functions, the random depth network is expanded beyond the constraints of a residual structure. The experimental results on the image classification datasets Cifar-10, Cifar-100 and Caltech-256 demonstrate that the proposed algorithm accelerates the convergence speed of the loss function and effectively improves the network's prediction accuracy. Additionally, it improves the training efficiency, alleviates overfitting, and achieves better generalization and accuracy. In most experiments, the T O P 1 accuracy of the model is increased by at least 2%, while the T O P 5 accuracy improves by at least 1%. The best result achieved by the proposed algorithm enhances the T O P 1 accuracy by 6.24%, and the T O P 5 accuracy by 3.62%. On the TGS Salt Identification and THUCNews datasets, the proposed algorithm still improved the performance of the model in image segmentation and text classification tasks, indicating that the proposed algorithm has good generalization ability.
On the one hand, nowadays, fake news articles are easily propagated through various online media platforms and have become a grand threat to the trustworthiness of information. On the other hand, our understanding of the language of fake news is still minimal. Incorporating hierarchical discourse-level structure of fake and real news articles is one crucial step toward a better understanding of how these articles are structured. Nevertheless, this has rarely been investigated in the fake news detection domain and faces tremendous challenges. First, existing methods for capturing discourse-level structure rely on annotated corpora which are not available for fake news datasets. Second, how to extract out useful information from such discovered structures is another challenge. To address these challenges, we propose Hierarchical Discourse-level Structure for Fake news detection. HDSF learns and constructs a discourse-level structure for fake/real news articles in an automated and data-driven manner. Moreover, we identify insightful structure-related properties, which can explain the discovered structures and boost our understating of fake news. Conducted experiments show the effectiveness of the proposed approach. Further structural analysis suggests that real and fake news present substantial differences in the hierarchical discourse-level structures.
No abstract is provided for this article.
This paper addresses the output feedback sliding-mode control (SMC) problem for discrete-time uncertain nonlinear systems through Takagi-Sugeno fuzzy dynamic models. Combining with the sliding surface, a descriptor system is constructed to characterize the sliding motion dynamics. Sufficient conditions for asymptotic stability analysis of the sliding motion are attained by the piecewise quadratic Lyapunov functions within a convex optimization setup. Two SMC design approaches are proposed to ensure the finite-time convergence of the sliding surface. Two simulation examples are presented to show the effectiveness of the proposed approaches.
This report considers attempts to develop dummy motorcyclists with breakable legs. Material characteristics are discussed. The variation in the scatter fracture load of different materials is compared using the Weibull modulus. The materials used in the different dummy legs have been calibrated statically and uni-axially whereas in crash tests multi-axial dynamic loads are sustained. The Independent Action criterion is used to show that: (1) compressive and torsion loads have only a small effect on bending; and (2) differences in results from different laboratories is the result of scatter in the material characteristics. The effect that leg fracture has on dummy trajectory is described using previously published experimental pedestrian impacts, motorcycle crash tests and pedestrian and car occupant computer simulation studies. Head trajectory is shown to be largely unaffected by leg fracture. For the covering abstract of the conference see IRRD 864606.
No abstract is provided for this article.
A simplified model predictive control algorithm is designed for discrete-time Markov jump systems with mixed uncertainties. The mixed uncertainties include model polytope uncertainty and partly unknown transition probability. The simplified algorithm involves finite steps. Firstly, in the previous steps, a simplified mode-dependent predictive controller is presented to drive the state to the neighbor area around the origin. Then the trajectory of states is driven as expected to the origin by the final-step mode-independent predictive controller. The computational burden is dramatically cut down and thus it costs less time but has the acceptable dynamic performance. Furthermore, the polyhedron invariant set is utilized to enlarge the initial feasible area. The numerical example is provided to illustrate the efficiency of the developed results.
Nowadays, each newly produced car must conform to the appropriate safety standards and norms. The most direct way to observe how a car behaves during a collision and to assess its crashworthiness is to perform a crash test. This paper deals with the wavelet-based performance analysis of the safety barrier for use in a full-scale test. The test involves a vehicle, a Ford Fiesta, which strikes the safety barrier at a prescribed angle and speed. The vehicle speed before the collision was measured. Vehicle accelerations in three directions at the centre of gravity were measured during the collision. The yaw rate was measured with a gyro meter. Using normal speed and high-speed video cameras, the behavior of the safety barrier and the test vehicle during the collision was recorded. Based upon the results obtained, the tested safety barrier, has proved to satisfy the requirements for an impact severity level. By taking into account the Haar wavelets, the property of integral operational matrix is utilized to find an algebraic representation form for calculate of wavelet coefficients of acceleration signals. It is shown that Haar wavelets can construct the acceleration signals well.
It is valuable for the real world to find the opinion leaders. Because different data sources usually have different characteristics, there does not exist a standard algorithm to find and detect the opinion leaders in different data sources. Every data source has its own structural characteristics, and also has its own detection algorithm to find the opinion leaders. Experimental results show the opinion leaders and theirs characteristics can be found among the comments from the Weibo social network of China, which is like Facebook or Twitter in USA.
In this paper by using the quantized space and time theory, the internal structure of the beauty quark has been analyzed, and the different modes for decay of beauty quark have been predicted theoretically. Also it has been explained that why the decay of beauty quark into kaon and muons is 88% of its decay into kaon, electron and positron.
Accepted version of an article from the journal: Nonlinear Dynamics and Systems Theory. Also available from the publisher: http://www.e-ndst.kiev.ua/v9n3.htm
In this paper a full-scale commercially available magnetorheological (MR) brake installed in a semi-active suspension (SAS) system is modeled and simulated. Two well-known phenomenological hysteresis models are explored: Bouc–Wen and Dahl ones. In particular, influence of their parameters on the response is evaluated and assessed. The next step is to introduce the artificial neural networks and discuss their application in the field of systems identification. Subsequently, two feedforward neural networks are created and trained to estimate parameters characterizing each of the MR damper models described. The semi-active suspension (SAS) system equipped with a MR brake is described and the detailed procedure for acquisition of the reference data used in the models validation stage is elaborated. The models outputs obtained by simulating them with the values of coefficients as identified by the networks are compared to each other as well as to the reference experimental data. Thanks to that, the comparative analysis between the suggested vibration suppression models and the full-scale MR brake is done and it is concluded which of the discussed models has a better performance. The usability of neural networks in the field of parameters estimation of the mathematical models of the real world phenomena is described as well. The novelty of the presented methodology is the application of artificial intelligence methods to estimate model parameters of a MR brake utilized in a SAS system. The results of this approach have a strong potential to be successfully implemented in the area of model-based control of semi-active vibration suppression systems.
This paper makes a prediction of Chinese stock index (CSI) future prices using fuzzy sets and multivariate fuzzy time series method. We select Chinese CSI 300 index futures as the research object. The fuzzy time series model combines the fuzzy theory and the time series theory, thus this model can solve the fuzzy data in stock index futures prices. This paper establishes a multivariate model and improves the accuracy of computation. By combing traditional fuzzy time series models and rough set method, we use fuzzy c-mean algorithm to make the data into discrete. Further more, we deal with the rules in mature modules of the rough set and then refine the rules using data mining algorithms. Finally, we use the CSI 300 index futures to test our model and make a prediction of the prices.
